A device and method for matrix transposition based on SRAM

Abstract

The invention discloses a device and a method for matrix transposition based on a static random access memory (SRAM). The device comprises an address decoding module, a read-write bus, a diagonal read control module and n matrix transposition storage bodies, wherein each matrix transposition storage body is formed by connecting n rows and n columns of matrix storage modules; a 14-pipe SRAM storage module is positioned on a diagonal from the left upper side to the right lower side, and a 12-pipe SRAM storage module is positioned outside the diagonal. The 14-pipe SRAM storage module accesses elements, which are positioned on the diagonal, in a matrix, and the 12-pipe SRAM storage module accesses elements, which are positioned outside the diagonal, in the matrix. According to the method, corresponding accesses for single words, single rows, single columns or single diagonals are executed according to the type of an input matrix and the difference of access modes, so that transposition is finished. The device and the method have the advantages of simple implementation method, simple and compact structure, low cost, high transposition speed, high efficiency, flexibility and multi-functionality.

Description

technical field [0001] The present invention relates to the field of matrix transposition data processing, in particular to a device and method for matrix transposition based on SRAM. Background technique [0002] Matrix Transposing is widely used in scientific computing, such as matrix decomposition, linear algebra solution, 3G communication algorithm, image video algorithm, radar, underwater acoustic algorithm and other scientific and engineering calculations. For an n×n order (n>1) source matrix A=[a ij ], each of its lower triangular elements a ij (i>j) along the main diagonal with its symmetric element a ji The exchange constitutes the transpose matrix A T [b ij ]. Equation (1) gives the transposed matrix A T [b ij ] and the source matrix A[a ij ] The corresponding relationship of each matrix element: [0003] b ji =a ij (i=0,1,...,n-1;j=0,1,...,n-1)(1) [0004] The application of fast and efficient matrix transposition methods and devices will greatly...

AI Technical Summary

Problems solved by technology

In the existing matrix transposition methods, whether it is through software, or through special hardware, or through SIMD (Single Instruction Multiple Data, Single Instruction Multiple Data) network structure, or using existing memory methods to realize the matrix transposition, cannot simply, efficiently, and flexibly meet the requirements of the matrix computer

Images